Are we There Yet? Immersive Virtual Reality to Improve Cognitive Function in Dementia and Mild Cognitive Impairment

Meelad Sayma, BMBS, BSc; Remco Tuijt, BA, MSc; Claudia Cooper, BMBS, MRCPsych, MSc; Kate Walters, BMBS, BMedSci, MRCGP, MMed, MSc, PhD


Gerontologist. 2020;60(7):e502-e512. 

In This Article


Principal Findings

We are unable to conclude whether the use of IVR for improving cognitive function in dementia, and MCI is effective. Some data suggest that IVR is acceptable to selected populations of people with dementia/MCI, with low attrition rates and some evidence suggesting people with dementia prefer and enjoy the use of virtual reality–based interventions (Maggio et al., 2018; Manera et al., 2016; Moyle et al., 2018; Optale et al., 2010). However, due to small sample sizes, it is important to highlight that these additional findings are not definitive either.

Despite this, there remains promise and incentives for future research, including well-designed randomized clinical trials with representative populations. The present studies in this field lay the groundwork and highlight opportunities to develop and rigorously test further interventions in this field.


Defining "Immersive Virtual Reality". A number of studies encountered during the search process utilized definitions of IVR that did not meet either criteria to be Virtual Reality, or were "non-immersive," as defined earlier in this article (see 2.6). This phenomenon is not unique to the field of dementia/MCI and not unique to health care (Garrett et al., 2018). In computer science, commercial spheres, and in the media, many define virtual reality as an ability to combine software with hardware to create a fully immersive experience (Rubin, 2018). However, the term "virtual reality" in health care has been used for over two decades to describe both nonimmersive and immersive experiences that create any version of an alternate reality (Riva, Bacchetta, Baruffi, Rinaldi, & Molinari, 1999). Although it is not possible to correct this disparity in definitions retrospectively, it is important to emphasize the need for future articles to utilize standardized definitions of IVR. Li and colleagues (2011) offer an extensive definition of immersive virtual reality through the hardware used to create this experience.

Garrett and colleagues (2018) highlight the importance of the concept of "presence" in virtual reality, referring to the extent to which users feel immersed in the world created by the components of software and hardware. Interestingly, presence was not addressed in our included studies. The level of presence is subjective, and dependent on user experience, and therefore may not be included in a definition of IVR; however, it is essential future studies assess the level presence experienced by users. This will allow for more direct comparison of IVR interventions, and to understand the direct additional benefit of the "immersive" experience virtual reality can offer over other simulated interventions. Assessment methods for presence have been developed such as the iGroup Presence Questionnaire (Scuhbert, Regenbrecht, & Friedmann, 2018). However, this questionnaire has not yet been validated for use with people with dementia or MCI (Scuhbert et al., 2018).

As described in the results, two studies utilized an HMD (Optale et al., 2010; White & Moussavi, 2016). Three studies utilized large screens or "video walls" as their visual hardware, two of these screens were considered "interactive" (Maggio et al., 2018; Manera et al., 2016; Moyle et al., 2018). As cheaper HMD options come onto the market, such as Google Cardboard, costing as little as 10USD to convert a smartphone into an HMD, it is important to highlight the somewhat prohibitive cost of utilizing video walls, costing over 10,000USD (Goh et al., 2018). Future studies should consider these barriers to access for IVR interventions when developing their own interventions (Ambron, Miller, Kuchenbecker, Buxbaum, & Coslett, 2018; Massetti et al., 2018).

Heterogeneity of Interventions. It is important to highlight the heterogeneity in hardware, software, and the specific types of training used in our included studies. This makes it difficult to draw specific conclusions about the benefits of IVR for people with dementia/MCI. It could be speculated that if a specific game is deemed effective or acceptable, it is unclear if a specific part of the hardware or the whole experience is what benefits the individuals. Quantifying this is difficult without control conditions for these factors, which three of the five included papers attempted to do by creating "paper" versions of each game delivered in virtual reality (Maggio et al., 2018; Manera et al., 2016; Optale et al., 2010). It is however more difficult to create a control condition for the software used in trials, as the immersive style of the experience may also influence the efficacy of the intervention. The lack of differentiation for content delivery is not a unique issue to the emerging field of IVR for people with dementia or MCI, but rather a problem seen throughout IVR health care developments (Garrett et al., 2018). However, this may mean that more than one control is needed to demonstrate the efficacy of IVR in a randomized control trial. One paper-based study simulating the content of the intervention and one computer-based study simulating the content of the intervention without immersive elements. This will allow for greater external validity and true demonstration of the benefits of IVR.

An additional challenge is presented by rapid development of IVR technology. Large differences in software and hardware availability exist, dramatically altering the experience of participants in studies even just 1 year apart (Faber, Patterson, & Bremer, 2013; Garrett et al., 2014; Parsons & Rizzo, 2008). This means it is fundamentally important for researchers to clearly document the year and version of each piece of hardware and software used, to allow for greater external validity—something that was not uniformly done well across the papers included in our review. We attempted to compare hardware by using a simple scoring system for the components of hardware used in the interventions. However, more detailed reporting of hardware components involved in IVR interventions would contribute to greater external validity.

Our review highlights a challenge specific to dementia and other complex neurological conditions. The domains affected by dementia and MCI are broad and just as multiple activities (in an MMSE or MoCA) are used to assess different domains, varying therapeutic activities may be needed to combat the degeneration of varying cognitive domains (Hodkinson, 2012; Creavin et al., 2016; Nasreddine et al., 2005). Only one of our five studies provided a clear rationale and breakdown of the activities in their intervention and their links and impact on specific cognitive domains, while an additional study focused on the cognitive domain of memory (Maggio et al., 2018; White & Moussavi, 2016). To improve external validity in future studies, it is important that a clear rationale behind specific games and interventions is explained, including which cognitive domains are targeted and how this might map onto real-life functioning and clinical outcomes for participants. The lack of theoretical basis of interventions also appears to be an issue in the wider literature surrounding clinical IVR, for example, when describing the mechanism of pain relief achieved by IVR (Garrett et al., 2018).

Acceptability and Ethics of True IVR. As discussed, full IVR involves the use of HMDs (Huygelier et al., 2019). Two issues arise in their use. First, some hypothesize that the use of HMDs may be generally unacceptable to the older population, with this technology being unfamiliar and potentially considered invasive. The second issue that arises in populations with more severe cognitive impairment is an issue of consent, especially if HMDs are considered an invasive use of technology. A recent study of 76 older, noncognitively impaired patients suggested that after first use, older people's attitudes toward HMDs and IVR changes from neutral to positive and reported minimal cyber sickness (Huygelier et al., 2019). The findings suggest that after first use, older people's attitudes toward HMDs and IVR changed from neutral to positive, and they reported minimal cyber sickness. The two studies that used HMDs with cognitively impaired populations were both included in this review (Optale and White). White and colleagues reported a case study, and Optale and colleagues used the HMD technology available in 2010, which may now be considered out of date. As such, no concise conclusions can be drawn, yet there seems to be an indication that it may be acceptable.

The ethical challenge of administering IVR who are unable to consent but may benefit from the technology is complex. The benefits from HMDs fully immersing an individual in a different environment may also be considered potentially harmful, removing patients from an environment that they may be more familiar with and potentially having unpredictable, disorientating, and distressing effects in severe cognitive impairment. In addition, the physical strapping of the HMD to an individual's head may be considered ethically unpalatable.

Challenges in Disseminating IVR Technology. One of the key barriers to conducting research with IVR technology and subsequent implementation is the costs involved. The current average cost of the top-selling VR headsets is £382 ($464) not including the additional computers or gaming devices needed to operate some devices (Greenwald, 2019). The costs associated with developing the software involved can be many multiples greater than the cost of the hardware. This may prohibit those attempting to develop and trial new interventions, especially in academic environments given the high rate of obsolescence of new technology (World Economic Forum, 2017). This may be overcome by forming new collaborations with games companies who have the resources to develop high-quality IVR interventions and the means to access consumers quickly once the interventions have been iterated and validated by academic centers. If IVR-based therapies are proven to be effective and useful for those with dementia attempts have to be made to mitigate any associated costs for patients.

Research Recommendations

Future researchers should:

  • ensure their interventions are fully immersive,

  • study the extent to which participants felt immersed in their environment,

  • utilize the latest IVR technology clearly document the version of technology used,

  • ensure adequate controls are present, to control for content of intervention and method of delivery (i.e., a paper-based control may not be enough to demonstrate that IVR is the superior method of delivery).

IVR studies should in addition be published without delay, to reduce risk of obsolescence by the time the study reaches mass readership.


Only a small number of studies with small samples and methodological limitations have been published using IVR in dementia. It is therefore difficult to draw conclusions other than that more evidence is needed to demonstrate the use, acceptability, and effectiveness of IVR in dementia and MCI. It is important to learn quickly from the gaps present in the published literature, highlighting a need for a universally accepted definition of immersive virtual reality, and more robust clinical trials, utilizing technology that is already available to consumers internationally.